Greinacher voltage multiplier

[EC8010]

Can anyone help? I'm looking for a copy of Heinrich Greinacher's 1920 paper on voltage multipliers that predates the Cockcroft/Walton patent. Greinacher was Swiss, so perhaps one of our Swiss colleagues?

'Erzeugung einer Gleichspannung vom vielfachen Betrage einer Wechselspannung ohne Transformator' Bulletin des Schweizer Elektrotechnischer Vereins 11 (1920)

[kevinkr]

Just so people know what you are talking about here is a useful link:

http://www.kronjaeger.com/hv/hv/src/mul/

Redrawn the Greinacher appears to be the standard "half wave" (operation actually is fullwave) doubler if you ignore the ground shown on the node between the caps. (It is described as a half wave multiplier in the literature, but in fact both ac polarities are used, but one polarity charges one cap, the other polarity the other cap. A question of semantics? I think here in the U.S. this is often referred to as a full wave doubler.) The "halfwave" Greinacher can also be looked at as two halfwave rectifiers referenced to ground, there is also a variant that today we would recognize as a center tapped bridge.

The Villard half wave doubler (see above comment) also mentioned on the site above can be cascaded for 2X or greater voltage multiplication. It looks suspiciously like the Cockcroft/Walton ladder multiplier unless I am missing something.

See here for Cockcroft/Walton: http://www.wenzel.com/pdffiles/voltmult.pdf

The multipliers are drawn differently on the two sites, but based on a quick inspection I believe the Villard and CW are basically the same thing.

[kevinkr]

Here's something on the Greinacher Cascade:

http://home.earthlink.net/~jimlux/hv/grein.htm

It would be very interesting to see the original 1920 paper particularly if available in an English translation.

[smoking-amp]

I noticed the voltage drop formula for the Villard cascade/ C-W multiplier is quite severe for large n (using the same size capacitors). Seems this could be mitigated by using larger capacitors toward the low voltage/high current end.

Ie, scale capacitor size to current carried at each point. Probably will only get a linear with n formula then for drop voltage.

Don

[Nordic]

Yeah, I use a small PCB transformer and a greinacher in my ecc88, works very well indeed.

[smoking-amp]

One can also duplicate the whole Villard/C-W multiplier network and use each such section with a different phase of AC drive to reduce the voltage drop and ripple. I have seen dental X-ray power supplies that used two cascades off a center tapped ferrite primary Xfmr to get 100KV.

But there is no real limit to the number of phases if using a S-S switch mode drivers setup, except each square wave input will then need a resonating inductor to smooth out the current to get sine wave like ripple. Four phases may then be ripple free altogether when using the resonators.

Don

[nhuwar]

I'm scratchen my head. Why would you want to use any sort of voltage multiplier in audio sense current draw causes substantial drops in output voltage?
Unless it's not for audio in that case never mind.

Nick

[SY]

If current draw is relatively constant, who cares?

Off the top of my head, the Marantz 9, H-K Citation, and Eico HF-87 are all excellent amps and all use a voltage doubler.

[nhuwar]

Wow that surprises me I would have never though that honestly, you learn something new every day.

Nick

[Eli Duttman]

Off the top of my head, the Marantz 9, H-K Citation, and Eico HF-87 are all excellent amps and all use a voltage doubler.

You bet! Good enough for Avery Fisher, Saul Marantz, and Stu Hegeman. Good enough for anyone. Properly executed, a "full wave" doubler B+ PSU is quiet and has a low impedance. We were able to get a GOOD B+ PSU at modest cost in "El Cheapo" by using a doubler.